UvA-DARE (Digital Academic Repository)

Contrasting nutritional strategies in two closely related marine mixotrophic chrysophytes

Wilken, S.; Choi, C.J.; Worden, A.Z. DOI 10.1111/jpy.12920 Publication date 2020 Document Version Final published version Published in Journal of phycology License CC BY Link to publication

Citation for published version (APA): Wilken, S., Choi, C. J., & Worden, A. Z. (2020). Contrasting nutritional strategies in two closely related marine mixotrophic chrysophytes. Journal of phycology, 56(1), 52-67. https://doi.org/10.1111/jpy.12920

General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Download date:28 Sep 2021 J. Phycol. 56, 52–67 (2020) © 2019 The Authors. Journal of Phycology published by Wiley Periodicals, Inc. on behalf of Phycological Society of America. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. DOI: 10.1111/jpy.12920

CONTRASTING MIXOTROPHIC LIFESTYLES REVEAL DIFFERENT ECOLOGICAL NICHES IN TWO CLOSELY RELATED MARINE PROTISTS1

Susanne Wilken2 Monterey Bay Aquarium Research Institute, 7700 Sandholdt Road, Moss Landing, California 95039, USA Department of Freshwater and Marine Ecology, Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, Science Park 904, Amsterdam 1098 XH, The Netherlands Chang Jae Choi, and Alexandra Z. Worden2 Monterey Bay Aquarium Research Institute, 7700 Sandholdt Road, Moss Landing, California 95039, USA Ocean EcoSystems Biology Unit, GEOMAR Helmholtz Centre for Ocean Research, Dusternbrooker€ Weg 20, Kiel 24105, Germany

Many marine microbial eukaryotes combine Key index words: chrysophytes; microbial food web; photosynthetic with phagotrophic nutrition, but mixotrophy; phagotrophy; phytoplankton incomplete understanding of such mixotrophic Abbreviations: AOX, mitochondrial alternative oxi- protists, their functional diversity, and underlying dase; ASW, artificial seawater; CIII, mitochondrial physiological mechanisms limits the assessment and respiratory complex III; ETR, electron transport modeling of their roles in present and future ocean rate; FALS, forward-angle light scatter; FISH, fluo- ecosystems. We developed an experimental system rescence in situ hybridization; FRRf, Fast repetition to study responses of mixotrophic protists to rate fluorometry; GFP, green fluorescent protein; availability of living prey and light, and used it to NPQ, non-photochemical quenching; PTOX, plastid characterize contrasting physiological strategies in terminal oxidase; VAZ, violaxanthin cycle pigments two stramenopiles in the genus Ochromonas.We show that oceanic isolate CCMP1393 is an obligate mixotroph, requiring both light and prey as Unicellular photosynthetic eukaryotes of diverse complementary resources. Interdependence of evolutionary origin have long been recognized as photosynthesis and heterotrophy in CCMP1393 globally important primary producers in the ocean comprises a significant role of mitochondrial (Field et al. 1998). As phytoplankton, these pig- respiration in photosynthetic electron transport. In mented protists are clearly distinguished from contrast, coastal isolate CCMP2951 is a facultative purely heterotrophic protists, which are considered mixotroph that can substitute photosynthesis by the major predators in marine microbial food webs, phagotrophy and hence grow purely recycling nutrients, and linking bacterial production heterotrophically in darkness. In contrast to to higher trophic levels (Pomeroy 1974, Azam et al. CCMP1393, CCMP2951 also exhibits a marked 1983). However, this strict dichotomy does not accu- photoprotection response that integrates non- rately reflect natural communities, which harbor photochemical quenching and mitochondrial many mixotrophic protists that combine the ability respiration as electron sink for photosynthetically to photosynthesize with phagotrophic consumption produced reducing equivalents. Facultative of microbial cells (Bird and Kalff 1986, Estep et al. mixotrophs similar to CCMP2951 might be well 1986, Flynn et al. 2013). Mixotrophs can be impor- adapted to variable environments, while obligate tant consumers of prokaryotic and eukaryotic mixotrophs similar to CCMP1393 appear capable of microbes (Havskum and Hansen 1997, Sanders resource efficient growth in oligotrophic ocean et al. 2000, Unrein et al. 2007, Hartmann et al. environments. Thus, the responses of these 2012, Orsi et al. 2018), but their concurrent contri- phylogenetically close protists to the availability of butions to carbon fixation via photosynthesis com- different resources reveals niche differentiation that plicate integration of their ecosystem roles into influences impacts in food webs and leads to biogeochemical models (Mitra et al. 2014, Worden opposing carbon cycle roles. et al. 2015). When both processes are integrated within the same cell, photosynthetic energy acquisi- 1 Received 14 February 2019. Accepted 13 August 2019. First Pub- tion can compensate respirational losses of ingested lished Online 16 September 2019. Published Online 1 November 2019, Wiley Online Library (wileyonlinelibrary.com). carbon (Stoecker and Michaels 1991) leading to 2Authors for correspondence: [email protected]; azworden@geo- higher carbon transfer efficiencies from prey to mar.de predator, which is predicted to result in more Editorial Responsibility: T. Mock (Associate Editor)

52 CONTRASTING MIXOTROPHIC LIFESTYLES 53 productive food chains (Ward and Follows 2016). specific group within the stramenopiles, the chryso- Concurrently, higher trophic transfer efficiencies phytes, have been identified as mixotrophic bacteri- result in lesser release of remineralized nutrients by vores in oligotrophic ocean regions (Hartmann mixotrophic compared to heterotrophic predators et al. 2013). The few studies that have evaluated (Rothhaupt 1997). However, the net impact of pha- chrysophyte abundances, either by enumeration of gotrophic mixotrophs on carbon cycling may vary plastid-containing chrysophytes using FISH (Jardil- widely between species or environmental conditions, lier et al. 2010) or by semi-quantitative dot-blot creating considerable uncertainty in how these hybridization (Lepere et al. 2009, Kirkham et al. organisms should be included in ecosystem models. 2013), identified them as a numerically important The accuracy of theory-based predictions that component of pico- and nanophytoplankton com- involve mixotrophy, such as trait-based approaches munities in the tropical North-East Atlantic, the (Andersen et al. 2015) or detailed physiological hyper-oligotrophic South Pacific gyre, the Arctic models (Flynn and Mitra 2009), is currently limited and Indian oceans. Chrysophyte sequences retrieved by the weak empirical basis for model assumptions from environmental assays are typically dominated and parameterization. by environmental clades (Lepere et al. 2009, del The combined capacity for photosynthesis and Campo and Massana 2011), and the roles of individ- phagocytosis in many extant eukaryotes is not sur- ual clades in the marine microbial food web remain prising. Phagocytosis is an ancient trait that has unknown. Chrysophytes include species with numer- been involved in acquiring photosynthetic cells as ous nutritional strategies from autotrophic to vari- endosymbionts, giving rise to the first eukaryotic ous mixotrophic and purely heterotrophic lifestyles, photosynthetic organelle, the plastid, and then with loss of photosynthetic capabilities having spreading into evolutionarily distant lineages of occurred multiple times independently (Boenigk eukaryotes through additional endosymbiosis events et al. 2005, Grossmann et al. 2016). However, physi- (Lane and Archibald 2008, Keeling et al. 2013, ological information for this group is largely based Worden et al. 2015). In some photosynthetic on cultured isolates from freshwater (Caron et al. eukaryotes, the capacity for phagocytosis has been 1990, Rothhaupt 1996), brackish, and coastal envi- lost, such as diatoms, which belong to the stra- ronments (Andersson et al. 1989, Floder€ et al. menopiles. However, the presence of mixotrophs 2006). It is unclear how far information from across distant branches of the eukaryotic tree (such coastal isolates or other evolutionarily distant groups as multiple independent lineages within the alveo- such as dinoflagellates can be extrapolated to ocea- lates, prymnesiophytes, and stramenopiles), with nic chrysophytes. The sole study of an oceanic chrys- different evolutionary histories, cellular morpholo- ophyte, Ochromonas CCMP1393, uses transcriptome gies, and metabolic potentials, suggests that plastid analyses to infer that it has a stronger dependence acquisition does not necessarily result in evolution on photosynthesis than a freshwater isolate (Lie toward specialist photoautotrophy (Selosse et al. et al. 2018). Mixotrophic strategies observed in 2017). Despite the phylogenetic diversity among freshwater and brackish Ochromonas species often mixotrophs their lifestyles can be classified func- show a strong heterotrophic component with the tionally into non-constitutive mixotrophs that ability to grow purely heterotrophically in darkness acquire their photosynthetic potential by hosting (Tittel et al. 2003, Palsson and Daniel 2004). A close photosynthetic endosymbionts or stealing plastids, interaction between photosynthetic and mitochon- and constitutive mixotrophs that possess their own drial electron flow as also known from diatoms plastids (Mitra et al. 2016). Additionally, mixo- (Allen et al. 2008, Bailleul et al. 2015) has been trophic strategies vary in the degree to which pho- hypothesized to underlie a tendency toward photo- toautotrophy and phago-heterotrophy contribute to heterotrophy in the freshwater O. danica (Wilken the overall nutrition (Stoecker 1998) and the phe- et al. 2014a). However, Ochromonas as defined today notypic plasticity that allows adjusting the strategy appears to be polyphyletic (Grossmann et al. 2016) in response to environmental conditions. Recent and the physiology of marine isolates has yet to be attempts have started to address this functional directly compared in feeding experiments. diversity of mixotrophs in models (Flynn and Mitra The nutritional diversity of chrysophytes provides 2009, Mitra et al. 2016, Berge et al. 2017, Leles a platform for investigating differentiated strategies et al. 2018), but underlying metabolic attributes in phylogenetically related organisms from various will need to be resolved to understand the biogeo- environments. The challenges to studying energetics chemical and ecological impact as well as evolu- and implications of different mixotrophic strategies tionary trajectories of photosynthetic eukaryotes. lie principally in the maintenance of predatory mix- The taxonomic identity of the plastid-containing otrophs in culture under axenic conditions and per- constitutive mixotrophs that often dominate bac- forming experiments with living prey. While terivory in marine environments is not well charac- mesocosm studies have employed bacteria express- terized and studies quantifying their impact rarely ing green fluorescent protein (GFP) as experimen- report community composition (but see Unrein tal amendments of traceable living prey (Worden et al. 2014). Multiple prymnesiophytes and a et al. 2006), this approach has yet to be employed 54 SUSANNE WILKEN ET AL. in culture-based studies. Here we developed a con- the antibiotic treatment was verified by testing for bacterial trolled experimental system to manipulate and track growth in a medium containing peptone and malt extract, and by epifluorescence microscopy based on staining with 15 nM the availability of live bacterial prey, using a marine 0 0 4 ,6-Diamidine-2 -phenylindole dihydrochloride (DAPI). Vibrio mutant that has a higher temperature opti- The non-motile FlrA-mutant of Vibrio fischeri ES114 (Milli- mum (Graf et al. 1994) than the eukaryotes under kan and Ruby 2003) was used as both a heat-killed bacterial study and harbors GFP on an antibiotic resistance prey source and a traceable living food source. Introduction plasmid. These systems were developed for two mar- of the pVSV102 plasmid further conferred kanamycin resis- ine chrysophytes expected to represent different tance and constitutive expression of GFP (Dunn et al. 2006). species, the coastal (CCMP2951) and oceanic Stock cultures of the mutant were maintained on LBS plates l 1 (CCMP1393) Ochromonas isolates. We compared the (Stabb et al. 2001) with 100 g mL kanamycin at room temperature. Prior to experiments, V. fischeri was transferred relative importance of photosynthesis and ingestion into liquid medium containing 100 lM phosphate, 821 lM of bacterial prey to nutrition of both isolates and ammonium chloride, 888 lM glucose, K-medium trace metal assessed the phenotypic plasticity in response to mix (Keller et al. 1987), 4 mM Tris HCl pH 7.5, and 100 lg variations in light intensity and prey availability. To mL 1 kanamycin in an ASW base. Bacteria were harvested in examine differences in their lifestyles more deeply, exponential phase (abundance between 0.8 and 1.3 9 109 1 ° we further tested for a potential interaction between cells mL ) and either heat-killed at 80 C for 30 min (for maintenance of chrysophyte stock cultures) or used as live nutritional pathways, by assessing the role of mito- bacterial prey for experiments. Bacterial cells were cen- chondrial respiration in photosynthetic electron trifuged at 10,000g for 12 min, washed in KASW medium, transport. Our study provides first insights into the and stored at 4°C overnight prior to use in experiments. underlying metabolic and physiological characteris- Flow cytometry. Cells were enumerated by flow cytometry tics of distinct nutritional strategies in predatory after fixation with freshly prepared, buffered formaldehyde mixotrophs from both coastal and oceanic environ- (pH 7.2; 1% final concentration) and dark incubation for ments. Collectively, such studies will improve our 20 min. Fixation in this manner did not induce visible prey ability to mechanistically integrate widespread mixo- egestion. Fixed samples were either run immediately on an Accuri C6 flow cytometer (BD Biosciences, San Jose, CA, trophic lifestyles into models used to predict carbon USA) or flash-frozen in liquid nitrogen and stored at 80°C cycling under future ocean conditions. until analysis on an Influx flow cytometer (BD Biosciences; for details see: Cuvelier et al. 2010). Red and yellow-green MATERIALS AND METHODS polystyrene fluorescent beads (Polysciences, Warrington, PA, USA) were added as internal standards. On the Accuri, the Chrysophyte isolates and growth conditions. The two Ochromo- detection threshold was set on forward-angle light scatter nas isolates, CCMP1393 and CCMP2951, were obtained from (FALS) for Ochromonas and on green fluorescence (510/ the National Center for Marine Algae and Microbiota 15 nm) for V. fischeri. For the latter, settings were validated (NCMA, East Boothbay, ME, USA). They originate from the against flow-cytometric bacterial counts after DNA staining North-West Atlantic Gulf Stream (38.70° N, 72.37° W) and a with SybrGreen (Invitrogen), to ensure that the GFP signal coastal site in the subtropical North-West Pacific (24.76° N, was sufficient to detect all bacteria. The chlorophyll-derived 125.44° E), respectively. Morphological differences indicate autofluorescence of Ochromonas cells was normalized to the that they might represent different species with an eyespot red beads and represents a proxy for chlorophyll content being present in CCMP2951 (as documented on the NCMA (Cuvelier et al. 2017). website: https://ncma.bigelow.org/ccmp2951#.XNvQzc DNA extraction, sequencing, and phylogenetic analy- TgpPY) but not CCMP1393. The cultures were grown on K- ses. CCMP2951 cells were harvested from 50 mL culture by medium containing both ammonium (50 lM) and nitrate centrifugation at 8,000g for 10 min. DNA was extracted using l (882 M) as nitrogen source and Na2-glycerophosphate the DNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA). Poly- (10 lM) as phosphorus source (Keller et al. 1987) in an arti- merase chain reaction (PCR) to amplify the 18S rRNA gene ficial seawater (ASW) base with a salinity of 34. This medium was performed using forward primer 50-ACCTGGTTGATCC was chosen because it had been developed specifically for TGCCAG-30 and reverse primer 50-TGATCCTTCYGCAGGTT oceanic microalgae and because the presence of ammonium CAC-30 (Moon-van der Staay et al. 2000) in a total volume of and glycerophosphate would prevent impaired photosynthetic 25 lL containing 0.2 lM of each primer, 1 lL template, and growth of Ochromonas due to potential inaccessibility of stan- 12.5 lL of HotStarTaq MasterMix (Qiagen). After initial acti- dard inorganic nutrient sources such as nitrate (Wilken et al. vation at 95°C for 15 min, 30 amplification cycles were per- 2014b). However, we note the possibility of the organic phos- formed consisting of 30 s at 95°C, 30 s at 50°C, 4 min at phorus source to also support low levels of osmotrophy. Cul- 72°C, followed by 10 min final extension at 72°C. PCR prod- tures received additions of heat-killed bacterial prey (see ucts were purified by electrophoresis, gel extracted (QIAquick below) and were maintained at 21°C on a 14:10 h light:dark Gel Extraction Kit; Qiagen), and cloned with the TOPO-TA cycle with a light intensity of 100 lmol photons m 2 s 1 cloning kit (Invitrogen, Carlsbad, CA, USA). Plasmids were unless indicated otherwise. Sanger-sequenced bi-directionally using plasmid primers Development of controlled prey experimental system. The two M13F/M13R and the two internal primers 502f and 1174r chrysophytes were rendered axenic by killing their undefined (Worden 2006). The resulting 1,786 bp CCMP2951 18S rRNA bacterial consortia using antibiotics. Specifically, they were gene sequence was deposited under accession MH420530 incubated with 8 lg mL 1 chloramphenicol for 24 h, trans- (NCBI NR). ferred into medium containing 40 lg mL 1 kanamycin, Near full-length 18S rRNA gene sequences representing 80 lg mL 1 neomycin, 100 lg mL 1 streptomycin, and chrysophyte diversity were retrieved from the SILVA database 150 lg mL 1 penicillin G, and grown in this antibiotic cock- (release 128, https://www.arb-silva.de/documentation/relea tail with transfer and feeding with heat-killed bacterial prey se-128/; Pruesse et al. 2007). These 126 sequences, 4 addi- (see below) every other day for a total of 2 weeks. Success of tional sequences retrieved from iMicrobe MMETSP database CONTRASTING MIXOTROPHIC LIFESTYLES 55

(Keeling et al. 2014), alongside the CCMP2951 sequence gen- continuously for four more days with daily sampling for flow erated herein and a Synchromophyceae sequence (as out- cytometry and dilution to 5 9 104 cells mL 1. Photo-physio- group) were aligned using MAFFT (Katoh and Standley logical sampling and measurements were performed after 2013) with default parameters. Gaps were masked using 3 d, when both isolates still exhibited positive growth rates Gblocks (Castresana 2000) and phylogenetic inferences made when receiving sufficient light (i.e., 100 lmol photons m 2 by maximum likelihood methods based on 1,444 homologous s 1), while afterwards CCMP2951 ceased growing in prey- positions implemented in RAxML (Stamatakis 2014) under deplete treatments. This experiment thus assessed the short- the gamma corrected GTR model of evolution with 1,000 term response to prey depletion rather than acclimated auto- bootstrap replicates, or PhyML 3.0.1 (Guindon et al. 2010) trophic growth, which the two isolates were not capable of. with the same substitution model and 100 bootstrap repli- Samples were taken for pigment analysis and measurements cates. Additional reconstructions were performed in MrBayes of both carbon fixation and prey ingestion rates. Fast repeti- 3.2.6 (Ronquist et al. 2012) and the final tree was produced tion rate fluorometry (FRRf) was used to record rapid light– using FigTree 1.4.0 (http://tree.bio.ed.ac.uk/software/figtree) response curves and the response of photosynthetic electron and RAxML topology. Literature searches were performed for transport to chemical inhibition of mitochondrial respiration all genera represented in the tree to retrieve information on or chlororespiration. their capabilities to perform photosynthesis and phagocytosis. Pigment analysis: For pigment analysis, 50 mL of culture We note that detection of phagotrophic capacity requires a tar- was filtered onto 25-mm GF/F filters (Whatman, GE Health- geted experiment while photosynthetic capacity is inferred care Life Sciences, Pittsburgh, PA, USA), frozen in liquid N2, from readily detectable pigmentation of cells. and stored at 80°C. Extraction and HPLC analysis were per- Experiments. Prior to experiments, the Ochromonas isolates formed at the analytical facility of Horn Point Laboratory at were acclimated to experimental conditions by maintenance in the University of Maryland Center for Environmental Science. semi-continuous culture for 10 generations with both light and Concentrations of chlorophyll a (Chla), fucoxanthin, carotene, prey being available. To achieve this, cell abundances were diadinoxanthin and the violaxanthin cycle pigments (VAZ) vio- kept between 5 and 10 9 104 cells mL 1 and live bacterial laxanthin (V), antheraxanthin (A), and zeaxanthin (Z) were prey was added daily to an abundance of 3 9 107 cells mL 1. quantified according to Van Heukelem and Thomas (2001). These abundances were established in preliminary feeding Carbon fixation: Net rates of carbon fixation were mea- experiments to avoid prey depletion to sub-saturating abun- sured as 14C-incorporation over 24 h. This incubation period 7 1 dances below 1 9 10 cells mL within 24 h (Rothhaupt was chosen because it reliably provides net rates of carbon 1996), but allow depletion of prey by three orders of magni- fixation, while shorter-term incubations can result in esti- tude over a longer period (3 d). Daily dilution of cultures with mates somewhere between gross and net rates, depending fresh medium and supplementation of live bacterial prey was on species and growth conditions (Milligan et al. 2015). For based on flow cytometry counts of both predator and prey. each of the triplicate cultures two 20 mL aliquots at 5 9 104 Three experiments were performed using the GFP traceable, cells mL 1 of Ochromonas received 250 lCi 14C-bicarbonate live bacterial prey. Experiment 1 addressed the growth response at the onset of the light period. Depending on the treat- to light intensity under mixotrophic conditions (with prey- ment, the cultures contained either 3 9 107 cells mL 1 of amendment), Experiment 2 addressed the growth response to Vibrio fischeri (prey-amended treatments) or no prey (prey- availability of light and prey, and Experiment 3 assessed physio- deplete treatments). One aliquot was incubated with light logical responses to light limitation and prey depletion. intensity and diel cycle matching acclimation conditions (15 For Experiment 1, light–response curves were performed in or 100 lmol photons m 2 s 1) and the second served as triplicate for both isolates under prey-amended conditions. dark control. After 24 h incubation, 100 lL samples from The light intensities used examined the same range of light 14C-incubations was added to 10 mL scintillation cocktail levels, with slight differences in specific levels between isolate (CytoScint; MP Biomedicals, Santa Ana, CA, USA) to mea- experiments, with CCMP1393 grown at 0, 15, 30, 80, 130, and sure the total 14C-activity. The remaining cultures from 14C- 210 lmol photons m 2 s 1 and CCMP2951 at 0, 20, 40, 57, incubations were filtered onto 25-mm GF/F filters to mea- 95, and 165 lmol photons m 2 s 1. Cells were enumerated sure activities in samples and dark control, respectively. Fil- daily by flow cytometry (Accuri C6) as described above. Biolog- ters were incubated in 5% HCl overnight to remove ical triplicates of each isolate were maintained for 10 genera- inorganic 14C. After receiving 10 mL of scintillation, cocktail tions and specific growth rates were then averaged over the samples were measured on a scintillation counter (Beckman last 5 d of the growth period for each biological replicate. Coulter LS 6500, Indianapolis, IN, USA). Rates of carbon Experiment 2 further assessed the potential for purely fixation were calculated according to Pennington and Cha- autotrophic or heterotrophic growth. To this end, triplicate vez (2000) and converted to cellular rates of carbon fixation cultures of both isolates were pre-grown mixotrophically with using the mean Ochromonas abundance accounting for daily prey-amendments (as described above) at a light inten- growth over the incubation time (Heinbokel 1978). sity of 100 lmol photons m 2 s 1, close to the optimum Ingestion rates: Ingestion rates were measured by prey dis- determined for each isolate (see results). Cultures were then appearance relative to a prey only control. For each strain split into (i) an autotrophic treatment that did not receive and light intensity, triplicate cultures were diluted to 5 9 104 additional prey; (ii) a mixotrophic treatment that continued cells mL 1 and Vibrio fischeri was added to a final abundance to receive daily prey-amendments; and (iii) a heterotrophic of 3 9 107 cells mL 1. Triplicate control flasks received the treatment that received daily prey-amendments, but was kept same abundance of V. fischeri and a volume of culture filtrate in darkness. Cultures were maintained for 7 d as described (0.2 lm) equal to the culture volume added to the grazing above, sampled daily, and enumerated by flow cytometry. treatment. Flow cytometry samples were taken after 0 and For Experiment 3, triplicate cultures of both isolates were 24 h of incubation and fixed, stored, and counted on the acclimated to limiting (low) light intensity of 15 or near-opti- Influx as described above. Ingestion rates were calculated mal (high) light intensity of 100 lmol photons m 2 s 1, according to Heinbokel (1978). For measurement of prey car- respectively, and maintained for 10 generations as described bon content, duplicate 20 mL aliquots of V. fischeri stock used above. Cultures were then split into a mixotrophic, prey- as bacterial prey were filtered onto pre-combusted GF-75 fil- amended treatment grown with continued addition of bacte- ters (Advantec, Dublin, CA, USA), stored at 20°C and dried rial prey as before and a prey-deplete treatment that did not at 50°C overnight prior to analysis at the analytical facility of receive additional prey. Both treatments were grown semi- Horn Point Laboratory at the University of Maryland, Center 56 SUSANNE WILKEN ET AL. for Environmental Science. Live V. fischeri used for the experi- and super-saturating light intensities relative to acclimation ments had a carbon content of 0.069 0.014 pg C cell 1. conditions. The effect of mitochondrial inhibitors on photo- Fluorescence kinetics of photosystem II (PSII): FRRf was used to synthetic electron transport is expressed as relative decrease D 0 measure PSII fluorescence kinetics. Rapid light–response in the effective quantum efficiency of PSII ( F/F m) com- curves were recorded using a sequence of eight light intensities pared to the control reading acquired before addition of from 0 to 300 lmol photons m 2 s 1. Cultures were inhibitors. exposed to the respective light intensity for 3 min before per- Statistics: To detect differences in growth between auto- forming four series of measurements at 20 s time intervals. For trophic, mixotrophic, and heterotrophic conditions over time dark acclimated readings (F0 and Fm), cultures were incubated growth curves were analyzed by repeated measures ANOVA. in darkness for 25 min prior to measurements. The excitation Pigment contents and carbon fixation were tested for differ- and relaxation protocols are described in Guo et al. (2018). ences between isolates as well as effects of light intensity and During excitation, fluorescence rises from a minimum (F0)or prey availability by three-way ANOVA. Effects of light and 0 steady-state (F s) value in dark-acclimated cultures or in actinic prey availability were also analyzed separately for each isolate 0 light, respectively, to a maximum (Fm or F m in darkness or by two-way ANOVA to assess their impact and interaction in actinic light, respectively) as reaction centers close during light each isolate individually. Because ingestion rates can only be saturation. Fluorescence parameters and the effective absorp- measured in prey-amended treatments, a two-way ANOVA was r 2 tion cross-section of PSII ( PSII in A ) were derived according used for the full data set, while a one-way ANOVA tested for to Kolber et al. (1998). Effective quantum efficiency of PSII the effect of light in each isolate individually. Carotene to D 0 = 0 0 0 was derived as F/F m (F m F s)/F m and maximum quan- Chla ratios were log-transformed to improve homoscedastic- tum efficiencies were calculated from dark acclimated readings ity. FRRf measurements of rapid light–response curves and = as Fv/Fm (Fm F0)/Fm. Rates of electron transport per func- inhibitor tests were analyzed by repeated measures ANOVA tional reaction center of PSII (ETRRCII) were estimated as in with light as repeated factor. Holm-Sidak tests were per- Schuback et al. (2015): formed for all post-hoc pairwise comparisons.  D = 0 F F m 3 ETRRCII ¼ E rPSII 6:022 10 RESULTS Fv=Fm

Phylogenetic tree of chrysophytes. The distribution of where E is irradiance and the factor 6.022 9 10 3 converts 2 2 nutritional strategies across the phylogenetic tree of lmol photons to photons and A to m . The exponential photosynthesis–irradiance curve (Webb et al. 1974) chrysophytes is complex, based on the current level of understanding and available isolates (Fig. 1). The  aE orders Hibberdiales and Synurales are composed of ¼ ETR ETR ETRmax 1 e max presumably purely photosynthetic organisms, while – Paraphysomonadida and Apoikiida contain solely was fitted to ETRRCII data from rapid light response curves using the iterative least-square nonlinear regression in Sigma-Plot 13. non-photosynthetic organisms. The two isolates Non-photochemical quenching (NPQ) was calculated as: studied here have 97% nucleotide identity between their 18S rRNA genes and belong to the Ochromon- ðÞ 0 ¼ Fm F m : adales, a group which contains both mixotrophic NPQ 0 F m and purely heterotrophic taxa. Moreover, the pres- ence of heterotrophic taxa in several Ochromon- CCMP1393 exhibited maximum fluorescence values (Fmax) adales subclades suggests photosynthetic machinery in low light instead of darkness and Fmax therefore replaced has been lost multiple times within this group. F in the calculation of NPQ. m Light–response curves (Experiment 1). The two iso- Inhibition of mitochondrial and chloro-respiration: The influ- ence of mitochondrial and chloro-respiration on electron lates of Ochromonas showed pronounced differences transport through PSII was assessed by FRRf measurements in in their responses to resource availability. When both the presence of inhibitors. Plastid terminal oxidase (PTOX), light and prey were available to support mixotrophic the electron acceptor in chlororespiration, was inhibited by nutrition CCMP1393 and CCMP2951 reached maxi- 1 mM propylgallate (Bailey et al. 2008). The mitochondrial mum growth rates of 0.58 0.04 and 0.82 0.04 alternative oxidase (AOX) and respiratory complex III (CIII) d 1, respectively (Fig. 2, a and b). Under these con- were inhibited by 1 mM salicylhydroxamic acid and 5 lM antimycin A, respectively (Bailleul et al. 2015). A combination ditions, differences in responses to light intensity of salicylhydroxamic acid and antimycin A was used to block were reflected in adjustments of cellular red fluores- mitochondrial respiration completely. Readings were taken cence levels. CCMP1393 showed increased red fluo- using prey-amended cultures pre-acclimated to 100 lmol rescence at low-light intensities (Fig. 2a; ANOVA: photons m 2 s 1 and subjected to 3 min of illumination F = 196.8, P < 0.001), while CCMP2951 showed l 2 1 4,10 by 20, 100, and 300 mol photons m s prior to the decreased fluorescence when shifting to a more measurement representative of sub-saturating, near-saturating,

FIG. 1. Phylogenetic reconstruction of chrysophytes based on 18S rRNA gene sequences. A total of 1,444 positions from 132 near full- length sequences were used. Tree topology is based on maximum likelihood inference and node statistical supports are indicated based on Bayesian posterior node probabilities, and percent of bootstrap replicates from two maximum likelihood methods (1,000 replicates in RAxML and 100 replicates in PhyML). Columns of colored squares indicate the nutritional capacities of described genera for photosynthe- sis (left column: green—known phototrophic potential, white—potential not reported in cultured isolates, grey—unknown potential) and for phagotrophic ingestion of prey (right column: black—known phagotrophic potential, white—potential not reported in cultured iso- lates, grey—unknown potential). [Color figure can be viewed at wileyonlinelibrary.com] CONTRASTING MIXOTROPHIC LIFESTYLES 57

EF165131 Uroglena americana CCMP2769 JF730783 uncult. eukaryote Uroglena AF123290 Uroglena americana CCMP1863 MMETSP1141 Ochromonadaceae sp. CCMP2298 EF165142 Ochromonas sp. CCMP1393 AJ236861 Spumella danica AY651079 Pedospumella sp. JBAS36 MH420530 Ochromonas sp. CCMP2951 KT905424 Ochromonas sp. MMDL4071 EF165136 Ochromonas distigma AC25 EF165125 Ochromonas sp. AC514 EF165137 Ochromonas sp. CCMP1147 Ochromonas EF165139 Ochromonas sp. CCMP1149 EF165124 Ochromonas sp. aestuarii AC24 U42382 Ochromonas sp. CCMP1278 JF794055 Dinobryon faculiferum RCC2290 JN934680 Dinobryon faculiferum RCC2293 D. faculiferum JF730841 uncult. eukaryote JF730782 uncult. eukaryote Env. clade EF165141 Dinobryon cf. sociale UTCC392 MMETSP0019 Dinobryon sp. UTEXLB2267 Dinobryon EF165140 Dinobryon cylindricum CCMP2766 FN429125 Ochromonas sp. 99X4 AB622325 uncult. freshw. eukaryote EF165126 Ochromonas sp. CCMP2761 AY651074 Poteriospumella lacustris JBM10 EF165110 Ochromonas sp. CCMP2767 EF165114 Poterioochromonas malhamensis SAG933.1c AY699607 Poterioochromonas sp. DO−2004 Poterioochromonas HM161745 Poterioochromonas sp. Y4 Ochromonadales EF165112 Ochromonas cf. gloeopara CCMP2718 EF165113 Ochromonas cf. gloeopara CCMP2060 EF165108 Ochromonas danica CCMP588 M32704 Ochromonas danica AY520447 Ochromonas sp. TCS−2004 EF165143 Ochromonas perlata CCMP2732 AF123294 Ochromonas sphaerocystis AY651077 Acrispumella msimbasiensis JBAF33 GU290068 uncult. eukaryote Acrispumella AJ236860 Spumella obliqua JQ967332 Spumella sp. CH3 Spumella DQ388565 Spumella−like flagellate 1036 DQ388551 Spumella sp. 194f AF123301 Epipyxis aurea CCMP385 AF123298 Epipyxis pulchra CCMP382 Epipyxis EF432525 Chrysophyceae sp. I76 EF165133 Ochromonas sp. CCMP1899 FN690692 uncult. stramenopile FM955256 Hydrurus foetidus AF123296 Phaeoplaca thallosa EF165145 Chrysocapsa paludosa

AF123283 Chrysocapsa vernalis Chrysocapsa Hydrurales EF165104 Chrysonebula flava M87331 Hibberdia magna EF165101 Chromulina cf. nebulosa CCMP2719 MMETSP1095 Chromulina nebulosa UTEXLB2642

AF123285 Chromulina nebulosa CCMP263 Chromulina Hibberdiales U42454 Oikomonas mutabilis AY520451 Oikomonas sp. SA−2.2 Oikomonas AF123287 Chrysamoeba mikrokonta CCMP1857 AF123286 Chrysamoeba pyrenoidifera CCMP1663 Chrysoamoeba EF165102 Chrysamoeba tenera UTCC273 GU290087 uncult. eukaryote AY919713 uncult. freshw. eukaryote Env. clade Chromulinales AB622339 uncult. freshw. eukaryote AB695484 uncult. eukaryote AB695535 uncult. eukaryote Env. clade AY520450 Oikomonas sp. SA−2.1 FR874548 uncult. mar. picoeuk. FR874750 uncult. mar. picoeuk. FR874438 uncult. mar. picoeuk. FR874370 uncult. mar. picoeuk. KJ762940 uncult. eukaryote KJ759296 uncult. eukaryote Env. clade FJ537348 uncult. mar. chrysophyte JF730768 uncult. eukaryote AY919791 uncult. freshw. eukaryote Env. clade U73225 Mallomonas adamas MUCC287 JQ955668 Mallomonas splendens CCMP1782 GU935632 Mallomonas punctifera GU935621 Mallomonas elongata GU935620 Mallomonas tonsurata Mallomonas GU935623 Mallomonas cratis MMETSP1167 Mallomonas sp. CCMP3275 GU935624 Mallomonas pseudocratis KM590566 Synura glabra EF165117 Synura petersenii KM590554 Synura conopea KM590559 Synura curtispina Synura KM590600 Synura spinosa Synurales KM590563 Synura echinulata KM590584 Synura mammillosa EF165119 Tessellaria volvocina CCMP1781 U73219 Tessellaria volvocina MUCC302 Tessellaria HF549063 Tessellaria lapponica AB022864 Paraphysomonas foraminifera JQ967303 Paraphysomonas sp. 23 JMS−2012 AF109325 Paraphysomonas vestita AB168053 Monas sp. TON−O AY651095 Spumella−like flagellate JBNZ43 JQ967322 Paraphysomonas sp. 12 JMS−2012 Paraphysomonas JQ967320 Paraphysomonas sp. 13 JMS−2012 KM516202 Paraphysomonas sp. SIOpierPara1 EF432518 Paraphysomonas imperforata JQ967331 Paraphysomonas sp. 21 JMS−2012 AY651090, Spumella−like flagellate JBC29 GQ863818 uncult. eukaryote

Nutrition KF130273 uncult. eukaryote Paraphysomonadida AB622322 uncult. freshw. eukaryote Env. clade Phototrophic AB622333 uncult. freshwater eukaryote Phagotrophic AF123300 Chrysosaccus sp. CCMP295 EF165120 Chrysosaccus sp. CCMP1156 Chrysosaccus Unknown M87332 Chromulina chionophila EF165107 Chromophyton cf. rosanoffii Statistical support AB474963 Chromophyton vischeri Chromophyton EF165146 Lagynion cf. ampullaceum CCMP2727 1/100/100 AF123288 Lagynion scherffelii CCMP465 Lagynion ≥ 0.98/90/90 FR865769 Ochromonas tuberculata CCAP933/27 AF123293 Ochromonas tuberculata CCMP1861 O. tuberculata ≥ 0.88/80/80 AY651098 Apoikiospumella mondseeiensis JBM08 ≥ 0.88/70/70 AY642705 uncult. eukaryotic picoplankton Apoikiospumella AY919829 uncult. freshw. eukaryote FJ971855 Apoikia sp. EK−2010a FJ971854 Apoikia sp. EK−2010a JYME2 Apoikia

KP404827 uncult. eukaryote Apoikiida FJ537340 uncult. mar. chrysophyte FJ537322 uncult. mar. chrysophyte FJ537315 uncult. mar. chrysophyte Env. clade AY919738 uncult. freshw. eukaryote AY919804 uncult. freshw. eukaryote AY919691 uncult. freshw. eukaryote 0.10 KF925343 Synchroma pusillum CCMP3072 58 SUSANNE WILKEN ET AL.

CCMP1393 CCMP2951 1.0 (a) (b) 3.0 2.5 ) 0.8 -1 2.0 0.6 1.5 0.4 1.0 Growth rate (d Growth 0.2 0.5

0.0 0.0 Chl. fluorescence units) (rel. 0 50 100 150 200 0 50 100 150 200 Light intensity (μmol photons m-2 s-1)

1.0 (c) (d)

0.8

) 0.6 -1

0.4

0.2 Growth rate (d Growth 0.0 * * ** * ** * ** * mixotrophic* ********** growth autotrophic growth -0.2 heterotrophic growth

1234567 1234567 Time (Days)

FIG. 2. Growth responses to light under prey-amended conditions in the two Ochromonas isolates. (a and b) Light–response curves and cellular chlorophyll a fluorescence of CCMP1393 and CCMP2951 during prey-amended growth in the light (mixotrophic growth; Experi- ment 1). Cultures were maintained semi-continuously with daily additions of Vibrio fischeri as prey. (c and d) Growth curves of CCMP1393 and CCMP2951 (Experiment 2) with daily prey-amendments in the light (mixotrophic growth), with daily prey-amendments in darkness (heterotrophic growth), or after ceasing prey-amendments in the light (autotrophic growth). Asterisk indicates significant difference from mixotrophic conditions (P < 0.05; Holm-Sidak test following a two-way repeated measures ANOVA). Error bars indicate standard deviation among biological triplicates. [Color figure can be viewed at wileyonlinelibrary.com] heterotrophic nutrition at low-light intensities similar among isolates, although CCMP2951 = < (Fig. 2b; ANOVA: F5,12 825.7, P 0.001). depleted its prey less strongly at low-light intensities CCMP1393 exhibited photoinhibition at 210 lmol (Figure S1 in the Supporting Information). In photons m 2 s 1, while CCMP2951 did not show CCMP1393, photosynthetic carbon fixation was signs of photoinhibition over the range of light levels nearly halved after prey had been depleted in the tested (maximum 180 lmol photons m 2 s 1). high-light treatment, while it remained unchanged Autotrophic and heterotrophic growth potential (Experi- in CCMP2951 (Fig. 3a; two-way ANOVA, Table 1). ment 2). When kept in darkness, CCMP1393 started For both isolates, the 7-fold decrease in light inten- to die after 3 d (Fig. 2c), while CCMP2951 could sity, from near-optimal to limiting conditions, signif- maintain growth rates of 0.40 0.03 d 1 by icantly reduced net rates of carbon fixation (Fig. 3a; heterotrophy with bacterial prey as the only source three-way ANOVA, effect of light, Table 1), of carbon. After daily prey-amendments had ceased, although their responses differed significantly CCMP1393 maintained relatively high growth rates (three-way ANOVA, light 9 isolate interaction, for a few days after depleting prey, whereas Table 1). While rates of carbon fixation decreased CCMP2951 showed an immediate drop in growth 12- to 20-fold in CCMP2951, they only decreased 4- rates (Fig. 2, c and d). However, after 7 d without to 6-fold in CCMP1393 (Fig. 3a). This difference is feeding CCMP1393 began to die while CCMP2951 in line with the increased chlorophyll content of did not exhibit measureable death rates. CCMP1393 at low-light intensities (Fig. 4a; two-way Response to light limitation and depletion of prey (Exper- ANOVA, Table 1). Ingestion rates showed a differ- iment 3). In line with these different resource ent pattern to cellular chlorophyll content with a requirements, the two isolates responded differently significant increase at higher light intensity in to light limitation and the depletion of prey over a CCMP1393 and decrease in CCMP2951 (Fig. 3b; period of 3 d. Time courses of prey depletion were two-way ANOVAs, Table 1). Hence, in prey- CONTRASTING MIXOTROPHIC LIFESTYLES 59 amended cultures carbon fixation and ingestion of CCMP1393 after prey depletion resulting in a covaried in response to light intensity in CCMP1393, population decrease, while at light-saturation the while they were inversely related in CCMP2951. decrease in growth after prey depletion (55%) was Although low prey and light availabilities both led less severe than in CCMP2951 (72%). Note that to reduced growth in both isolates (Fig. 3c; ingestion rates represent gross carbon uptake from P < 0.001, Table 1), their responses to these prey, while carbon fixation is represented by net resources differed significantly (two-way ANOVA rates. Therefore, measured rates of carbon acquisi- interaction terms, Table 1). CCMP1393 grew gener- tion cannot be directly linked to growth responses. ally faster in high-light conditions, while CCMP2951 A decrease in FALS upon prey depletion (Figure S2 always showed higher growth rates when prey- in the Supporting Information) indicates decreased amended. Overall, the larger dependency on photo- cell sizes, which might further skew the relationship synthesis in CCMP1393 resulted in a relatively stron- between cellular carbon acquisition and specific ger reduction of growth under low-light conditions growth rates. (67%) compared to CCMP2951 (37%). Under low- Changes in pigmentation in response to light and light carbon fixation was not sufficient for survival prey availability also differed between isolates (Fig. 4; ANOVA interaction terms, Table 1). Cellu- lar Chla content increased significantly under low- CCMP1393 CCMP2951 light compared to high-light conditions in 25 (a) CCMP1393, but not in CCMP2951 (Fig. 4a; two-way ANOVAs, Table 1), as already indicated by red fluo-

) prey-amended

-1 20 prey-deplete rescence. Prey depletion was associated with

· d decreases in cellular chlorophyll content in -1 15 CCMP1393, but there was no significant change in CCMP2951 (Table 1). This pattern is in line with 10 the stronger reduction in FALS and therefore possi-

Net C-fixation bly cell size observed in CCMP1393 upon prey

(pg C · cell (pg C 5 depletion (Fig. S2). The adjustment of accessory pigments relative to Chla was more comparable 0 between isolates with fucoxanthin being slightly (b) increased at low-light intensities (three-way ANOVA,

) 20 Table 1). The photo-protective pigment carotene -1 was present at higher concentrations in CCMP2951

· d and increased at high-light intensities in both iso- -1 15 lates (three-way ANOVA, Table 1). The same was 10 true for diadinoxanthin which was present at 13-fold Ingestion higher concentrations in CCMP2951 compared to CCMP1393 under prey-amended conditions when

(pg C · cell · (pg C 5 acclimated to near-optimal high-light intensity (Fig- 0 ure S3 in the Supporting Information). In low-light 1.0 acclimated CCMP1393, diadinoxanthin was not (c) detectable. The de-epoxidized form diatoxanthin, 0.8 that is part of the diadinoxanthin cycle in diatoms 0.6 (Arsalane et al. 1994), was not detected in either Ochromonas isolate. The ratio of violaxanthin cycle

) 0.4

-1 pigments (VAZ-pigments) to Chl a was comparable

(d 0.2 among isolates and elevated under high-light accli- mation (Fig. 4d; three-way ANOVA, Table 1). In Growth rate Growth 0.0 CCMP2951, it was significantly higher during prey- -0.2 amended compared to prey-deplete conditions, -0.4 while the opposite trend was observed in CCMP1393 15 100 15 100 (two-way ANOVA, Table 1). Light intensity (μmol photons · m-2 · s-1) Maximum ETRRCII (ETRmax) was reduced in low- light compared to near-optimal, high-light accli-

FIG. 3. Carbon acquisition and growth of the two Ochromonas mated cultures (Fig. 5, a and b, Table 2) and this isolates under prey-amended and deplete conditions. Rates of (a) effect was strongest in CCMP2951. ETRRCII did not net primary production, (b) ingestion, and (c) growth in show photoinhibition in response to short-term CCMP1393 and CCMP2951 acclimated to 15 (LL) or 100 (HL) exposure to super-saturating light intensities, even lmol photons m 2 s 1 under either prey-amended or prey- deplete conditions. Error bars indicate standard deviation among though photoinhibition had been observed in accli- biological triplicates. For figure simplicity, results of statistical mated growth rates of CCMP1393 (Fig. 2a). The comparisons are shown in Table 1. capacity for non-photochemical quenching (NPQ) 60 SUSANNE WILKEN ET AL.

TABLE 1. Statistical analysis of results from Experiment 3. Effects of Ochromonas isolate phylogenetic identity, light intensity, and prey availability as well as their interactions in a three-way ANOVA and effects of light intensity, and prey availability as well as their interaction in each isolate individually by a two-way ANOVA. Shown are effects on carbon fixation (Cfix), rates of ingestion and growth, as well as Chla content, and relative contents of carotene (Caro), fucoxanthin (Fuco), and violaxanthin cycle pigments (VAZ). Arrows indicate the direction of significant responses with increasing light intensity or prey availability, respectively.

Three-way ANOVA Two-way ANOVA Single effect Isolate Interaction Parameter Factor df1 df2 FP F PCCMP1393 CCMP2951 Cfix/ cell Isolate 1 16 26.6 <0.001 Cfix/ cell Light 1 16 2668 <0.001 204.2 <0.001 ↑a ↑a Cfix/ cell Prey 1 16 0.64 0.436 83.0 <0.001 ↑a -a Ingestion Isolate 1 8 8.01 0.022 Ingestion Light 1 8 0.44 0.526 44.8 <0.001 ↑↓a Growth Isolate 1 16 93.9 <0.001 Growth Light 1 16 309.9 <0.001 17.2 0.001 ↑a ↑a Growth Prey 1 16 540.3 <0.001 26.2 <0.001 ↑a ↑a Chl a/cell Isolate 1 16 1.58 0.227 Chl a/cell Light 1 16 6.01 0.025 34.6 <0.001 ↓a -a Chl a/cell Prey 1 16 17.3 <0.001 36.5 <0.001 ↑a -a Caro/Chl a Isolate 1 16 6810 <0.001 Caro/Chl a Light 1 16 2742 <0.001 427 <0.001 ↑a ↑a Caro/Chl a Prey 1 16 30.2 <0.001 0.70 0.416 ↓a ↓a Fuco/Chl a Isolate 1 16 363 <0.001 Fuco/Chl a Light 1 16 157 <0.001 1.58 0.226 ↓↓a Fuco/Chl a Prey 1 16 5.81 0.028 8.94 0.009 ↓ -a VAZ/Chl a Isolate 1 16 963 <0.001 VAZ/Chl a Light 1 16 2205 <0.001 96.2 <0.001 ↑a ↑ VAZ/Chl a Prey 1 16 0.03 0.865 82.7 <0.001 ↓a ↑ aSignificant interaction of prey and light availability (P < 0.05). ·

prey-amended

prey-deplete ·

μ ·

FIG. 4. Pigmentation of the two Ochromonas isolates under prey-amended and prey-deplete conditions. (a) Cellular chlorophyll content and relative content of (b) carotene, (c) fucoxanthin, and (c) Violaxanthin cycle pigments (VAZ) in CCMP1393 and CCMP2951 accli- mated to 15 (LL) or 100 (HL) lmol photons m 2 s 1 under either prey-amended or prey-deplete conditions. Relative pigment contents are expressed as molar ratios to Chl a. Error bars indicate standard deviation among biological triplicates. For figure simplicity, results of statistical comparisons are shown in Table 1. CONTRASTING MIXOTROPHIC LIFESTYLES 61

CCMP1393 CCMP2951

) 300 (a) -1 (b)

· s 250 -1 200 · RC - 150 (e 100 RCII 50 ETR 0 (c) (d) 2.0 HL pre-acclimated, prey-amended HL pre-acclimated, prey-deplete 1.5 LL pre-acclimated, prey-amended LL pre-acclimated, prey-deplete

NPQ 1.0

0.5

0.0 100 (e) (f) Inhibition of: AOX CIII + AOX (%) CIII PTOX m 80

F/F' 60 Δ 40 20 Inhibition 0 0 100 200 300 0 100 200 300 Light intensity (μmol photons · m-2 · s-1)

FIG. 5. Electron transport in the two Ochromonas isolates under prey-amended and deplete conditions. (a and b) Rapid light–response curves of electron transport through photosystem II. Lines represent photosynthesis–irradiance curves fitted to the data (see Table 2 for parameter values). (c and d) Non-photochemical quenching (NPQ) in CCMP1393 and CCMP2951 pre-acclimated to low-light (15 lmol pho- tons m 2 s 1) and high-light (100 lmol photons m 2 s 1) intensities under either prey-amended conditions or after depletion of prey for 3 d. (e and f) Effect of inhibition of mitochondrial respiratory complex III (CIII), mitochondrial alternative oxidase (AOX), both CIII and AOX, or the plastid terminal oxidase (PTOX) involved in chlororespiration on effective photochemical yield of PSII in CCMP1393 and CCMP2951. Measurements were performed with mixotrophic, prey-amended cultures acclimated to 100 lmol photons m 2 s 1 and exposed to a sub-saturating light intensity, the acclimation intensity, or a super-saturating intensity for a period of 3 min prior to measure- ments. Error bars indicate standard deviation among biological triplicates. [Color figure can be viewed at wileyonlinelibrary.com] differed largely between isolates (Fig 5 c and d). modified the effective quantum yield of PSII under CCMP1393 exhibited only modest levels of NPQ, exposure to increasing light intensities (Fig. 5, e and = < while CCMP2951 showed strong NPQ upon expo- f; RM-ANOVA: F2,1 697, P 0.001 for CCMP1393 = < sure to high-light intensities, especially in the cul- and F2,1 491, P 0.001 for CCMP2951). This indi- tures that had been pre-acclimated to low-light cates both pathways act as an overflow mechanism for intensities. Additionally, for low-light acclimated reducing equivalents produced by the photosynthetic CCMP2951, the NPQ response was stronger in prey- machinery. In CCMP1393, photosynthetic yield was amended compared to prey-deplete cultures (Holm- affected by inhibition of CIII and AOX both individu- Sidak test, P < 0.02 for actinic light >100 lmol pho- ally and in combination. Inhibition of the entire tons m 2 s 1). mitochondrial respiration already affected photosyn- To assess alternative routes of electron transport thetic electron transport at sub-saturating light inten- and their role in photosynthesis, three specific inhibi- sities (Holm-Sidak test, P < 0.001 for combination of tors were used, targeting PTOX involved in chlorores- inhibitors compared to each individual one at l 2 1 piration or the mitochondrial electron acceptors CIII 20 mol photons m s ) and this effect did not and AOX, respectively. In both isolates, inhibition of increase strongly moving into saturating and super- mitochondrial- or chloro-respiration increasingly saturating intensities (100 and 300 lmol photons 62 SUSANNE WILKEN ET AL.

ABLE – T 2. Parameters of photosynthesis-irradiance curve fits to rapid light response curves of the two Ochromonas isolates based on electron transport rates (Fig. 5, a and b). Light intensities are given in lmol photons m 2 s 1. Values in parentheses give estimates of standard errors.

ɑ 2 Isolate Light Prey ETRmax adj r CCMP1393 100 amended 289 (8) 1.42 (0.07) 0.998 CCMP1393 100 deplete 370 (10) 1.37 (0.05) 0.999 CCMP1393 15 amended 207 (5) 1.30 (0.07) 0.996 CCMP1393 15 deplete 232 (5) 1.35 (0.05) 0.997 CCMP2951 100 amended 303 (8) 1.12 (0.04) 0.999 CCMP2951 100 deplete 271 (8) 1.19 (0.06) 0.998 CCMP2951 15 amended 123 (1) 1.33 (0.04) 0.997 CCMP2951 15 deplete 133 (2) 1.52 (0.06) 0.996

m 2 s 1, respectively). In CCMP2951, photosynthe- for instance, be much lower as reported in two sis was affected by inhibition of AOX but not by inhi- green algal Micromonas species (Worden et al. bition of CIII. The simultaneous inhibition of both 2009). Overall, our results exemplify the difficulties mitochondrial electron acceptors had no significant in linking phylogenetic to functional diversity of effect at sub-saturating light intensities, but caused a mixotrophs (Figs. 1 and 6). stronger effect compared to inhibition of AOX alone, Organisms with a mixotrophic lifestyle face a at higher light intensities (Holm-Sidak test, P < 0.001 trade-off between investments into nutrient uptake, for combination of inhibitors compared to each indi- photosynthesis, and phagotrophy (Andersen et al. vidual one at 100 and 300 lmol photons m 2 s 1). 2015). Modeling studies suggest that differential investment into optimization of these traits under- lies successional patterns of phytoplankton (Berge DISCUSSION et al. 2017). Similarly, the high phenotypic plasticity The two mixotrophic chrysophytes studied here in the facultative mixotroph CCMP2951 observed showed opposite physiological responses to environ- here reflects differential investment into these traits mental drivers. The capability of CCMP2951 to that is shaped by environmental conditions. At low- adjust its nutrition from pure heterotrophy in dark- light intensities, increased ingestion rates substitute ness to mixotrophy in the light and at least survival for photosynthesis and make additional investment by autotrophy in the absence of prey characterizes it into chlorophyll biosynthesis obsolete. The capacity as a facultative mixotroph, albeit tending more of phagotrophy to sustain growth in darkness is in toward the heterotrophic side of the mixotrophic agreement with previous studies of predominantly spectrum. Overall, the phenotypic plasticity in its heterotrophic freshwater isolates of Ochromonas and nutrition should facilitate growth under a wide Poterioochromonas (Andersson et al. 1989, Sanders range of environmental conditions. Similar to et al. 1990, Wilken et al. 2013), although its maxi- CCMP2951, CCMP1393 shows very poor autotrophic mum growth and ingestion rates observed here are growth in the absence of prey, although it is able to lower than reported for freshwater Ochromonas iso- maintain growth for a few days without feeding lates (Rothhaupt 1996, Wilken et al. 2013). More- (Fig. 2). It might thus have a slightly lower require- over, the high nutritional plasticity requires that the ment for prey and tend more toward the auto- two nutritional pathways can principally function trophic side of the mixotrophic spectrum compared independently from each other, as both can supply to CCMP2951. CCMP1393 is furthermore unable to sufficient resources to fuel cellular metabolism. In grow heterotrophically in darkness, and hence is an particular, under mixotrophic growth at low-light obligate mixotroph that requires both light and compared to high-light intensities, higher rates of prey for growth, at least under the conditions tested ingestion seem to compensate for the decreased herein. These different physiologies appear to be rates of carbon fixation resulting in an inverse rela- rooted in the interaction between photosynthetic tionship between these two rates (Fig. 3). Thus, our and heterotrophic processes, resulting in distinct results demonstrate that photosynthesis and ecological strategies for these two chrysophytes. The heterotrophy act as partly substitutable routes of two isolates thus likely represent different species, as resource acquisition in the facultative mixotroph. also supported by morphological differences with an Unlike facultative mixotrophy, photosynthesis and eyespot present in only CCMP2951 and detectable heterotrophy covary in the obligate mixotroph differences in their 18S rRNA genes. While their CCMP1393 (Fig. 3) and thus constitute complimen- 97% nucleotide identity in 18S rRNA genes indi- tary routes of resource acquisition that cannot func- cates close relatedness, this gene is strongly con- tion independently from each other. Such a strategy served and does not always provide sufficient cannot be explained solely by differential invest- resolution to distinguish closely related species. The ment into resource acquisition traits. Instead, the percentage of shared genes in their genomes might, reliance on the simultaneous performance of both CONTRASTING MIXOTROPHIC LIFESTYLES 63

Low light High light High light Prey available Prey available Prey depleted

h C C CI I C I CIII CIII p AOX III AOX AOX

o

r TCA TCA TCA

3

t

9

o

3

x

i

1

P

m PS II PS I PS II PS I PS II PS I

e M PTOX PTOX PTOX t CBC CBC CBC

C

a

C

g

i

l

b

o

h

p CI CI C CIII CIII I CIII o AOX AOX AOX

r

t TCA TCA TCA

1

o

5

x

i

9

2

m

P e PS II PS I PS II PS I PS II PS I

v M PTOX CBC PTOX PTOX

i

t C CBC CBC

a

C

t

l

u

c

a

f

FIG. 6. Responses of the two mixotrophic Ochromonas isolates to different environmental conditions. Both isolates reach highest growth rates under saturating light and prey availability. The obligate mixotroph (CCMP1393, top row) has higher chlorophyll content under low- light compared to high-light conditions and when prey is available compared to prey-deplete conditions. Rates of both carbon fixation in the Calvin–Benson cycle (CBC) and ingestion of prey are highest under mixotrophic conditions under high light. Reducing equivalents are exchanged between the plastid and mitochondria linking the photosynthetic with the respirational electron transport chain. Growth is most strongly reduced at low-light intensities. In contrast, the facultative mixotroph (CCMP2951, bottom row) increases its chlorophyll content when prey is depleted at high light, while in the presence of prey light has little impact on chlorophyll content. Rates of ingestion are highest under low-light condition and maximum rates of carbon fixation are reached when prey is depleted. Export of reducing equivalents from the chloroplast to mitochondria only acts as overflow during high-light condition. Growth is most strongly reduced in the absence of prey. AOX, alternative oxidase; TCA, tricarboxylic acid cycle; CI, respiratory complex I; CIII, respiratory complex III; PSII, photosystem II; PTOX, plastid terminal oxidase; PSI, photosystem I; CBC, Calvin–Benson cycle. [Color figure can be viewed at wileyonlinelibrary.com] photosynthesis and phagotrophy to attain positive reported in the chrysophyte Uroglena americana, growth rates suggests an integration of these two which appears to be more closely related to nutritional pathways that restrict their independent CCMP1393 than CCMP2951, although support in execution. This interdependence of photosynthesis this region of the tree is largely lacking (Fig. 1), and heterotrophy results in decreased ingestion and the dependency was further characterized as rates under light limitation and decreased chloro- being a requirement for bacterially produced phos- phyll content when prey are lacking, in line with pholipids (Kimura and Ishida 1989). A recent study changes in transcriptional patterns observed in this on CCMP1393 did not detect transcripts that isolate (Lie et al. 2018). Dependency on light for encode nitrate transporters, and concluded that this survival has been described in other mixotrophs chrysophyte does not have the capacity to utilize such as the brackish chrysophyte Ochromonas minima nitrate (Lie et al. 2018). Additionally, based on (Floder€ et al. 2006) and several dinoflagellates growth experiments the freshwater chrysophyte (Jeong et al. 2010, Hansen 2011). The increased Ochromonas globosa relies on reduced nitrogen chlorophyll content in CCMP1393 under low-light sources (Wilken et al. 2014b). However, in our conditions (Figs. 2a and 4a) reflects a photoacclima- experiments, both ammonium and nitrate were pro- tion response typical for photoautotrophs that vided which would presumably obviate reliance on results in greater light-harvesting capacity under low prey as nitrogen source depending on ammonium light to help maintain photosynthesis (MacIntyre depletion rates, which can be rapid in algal culture et al. 2002). Despite the potential for photosyn- experiments (McDonald et al. 2010). In general, thetic carbon acquisition, this route is not sufficient biosynthetic pathways required for purely photoau- to support stable positive growth rates in CCMP1393 totrophic growth could have been lost by some con- under the conditions tested herein (Fig. 2c). A stitutive, obligate mixotrophs, although in the dependence on phagotrophy is usually explained as absence of complete genome sequences the preva- reflecting a mechanism for acquiring essential lence of such pathway reductions remains an open growth factors or nutrients from prey. For example, question. increasing rates of phagotrophy have been observed An obligate requirement for phagotrophy in pho- with increasing light intensity in the freshwater tosynthetic eukaryotes can be expected to result in a chrysophyte Dinobryon cylindricum, and attributed to finely tuned integration of photosynthetic and het- providing necessary nutrients via prey cell digestion erotrophic metabolisms, regardless of whether this (Caron et al. 1993). Similar dependencies were is the cause or consequence of an obligately 64 SUSANNE WILKEN ET AL. mixotrophic lifestyle. A close energetic coupling light harvesting related protein (LHCSR and LHX) between plastids and mitochondria involving an found in diatoms and green algae (Peers et al. overflow mechanism for excess reducing equiva- 2009, Bailleul et al. 2010, Guo et al. 2018). In addi- lents, as well as support of carbon fixation by mito- tion, the formation of a pH gradient across the thy- chondrially derived ATP, also occurs in diatoms lakoid membrane is required for the induction of (Allen et al. 2008, Bailleul et al. 2015). Effects of NPQ. In CCMP1393, a slight NPQ response is mitochondrial inhibitors must generally be inter- already present in darkness, indicative of a pH gra- preted with care, as indirect effects are difficult to dient formed through chlororespiration (Jakob exclude. Nevertheless, in CCMP1393, photosynthetic et al. 1999). If the close interaction between plastids electron transport seems to be strongly influenced and mitochondria in this isolate supports a pH gra- by mitochondrial respiration (Fig. 6). In addition to dient across the thylakoid membrane independently its role as an overflow mechanism for reducing of light, a strong NPQ response might not be bene- equivalents observed in both isolates, mitochondrial ficial as it could restrict photosynthesis at low-light respiration acts as an electron sink even under sub- intensities. saturating light conditions in CCMP1393. It there- The two mixotrophic strategies described here fore likely represents an integral part of photosyn- likely reflect adaptation to different niches, thetic electron flow in this obligate mixotroph, although the specific competitive advantages gained involving both AOX and CIII, which might be used within complex communities is difficult to assess for ATP production from photosynthetically derived and the environmental distributions of mixotrophic electrons. While this could potentially reduce the strategies remain largely unstudied. A first assess- need for heterotrophic respiration and hence ment of the biogeography of mixotrophs revealed a increase carbon assimilation efficiencies from prey, gap in information regarding non-dinoflagellate higher rates of carbon fixation in the presence of constitutive mixotrophs (Leles et al. 2019), such as prey do not indicate a tendency toward photo- the chrysophytes studied here. In modeling studies, heterotrophy as suggested for the freshwater chryso- constitutive mixotrophs are typically considered to phyte Ochromonas danica (Wilken et al. 2014a). In be primarily autotrophic, supplementing their nutri- the facultative mixotroph, CCMP2951 mitochondrial tion by ingestion of prey if either light or nutrients respiration was negligible as an electron sink under become limiting and thus have an advantage under low-light intensities and the overflow mechanism at nutrient limitation (Leles et al. 2018). This is a strat- higher light intensities mainly relied on alternative egy, for example, found in prymnesiophytes and oxidase, which does not contribute to the proton some dinoflagellates (Hansen and Hjorth 2002, gradient for ATP production (Vanlerberghe and Hansen 2011). However, the spectrum of constitu- McIntosh 1997). Hence, in contrast to the obligate tive mixotrophs, ranging from preferentially mixotroph, mitochondrial respiration is not heterotrophic to obligately mixotrophic is less com- required as an integral part of photosynthetic elec- monly addressed. In a similar manner to which pri- tron transport in the facultative mixotroph. marily autotrophic mixotrophs have an advantage Differences in phenotypic plasticity between the competing for nutrients against non-phagocytotic two isolates studied here are further reflected in photoautotrophs in nutrient limiting conditions, pri- their responses to super-saturating irradiances. In marily heterotrophic mixotrophs could have an the obligate mixotroph, growth rates were reduced advantage over purely heterotrophic taxa at low prey under long-term exposure to high light despite the availability, if sufficient light is available (Tittel et al. maintenance of high electron transport rates during 2003, Fischer et al. 2017). Both types of mixo- short-term exposure. This indicates a downstream trophic strategies could thus be advantageous in limitation to electron transport that could not be oligotrophic environments. Obligate mixotrophy counteracted by photoacclimation. The lower con- has been implemented in modeling studies in the tent of the photoprotective pigments carotene, and form of kleptoplasty, where the capacity of a hetero- diadinoxanthin in the obligate mixotroph, relative trophic protist to photosynthesize depends on its to the facultative mixotroph, might contribute to its ingestion of autotrophic prey (Mitra et al. 2016). higher susceptibility to photoinhibition. The light- Although based on different underlying physiology, induced de-epoxidized form of diadinoxanthin, dia- kleptoplastic protists dominated predicted plankton toxanthin, is responsible for some forms of NPQ in biomass in a nutrient-limited scenario (Leles et al. other stramenopiles, specifically diatoms (Arsalane 2018), hinting at the success of obligate mixotrophs et al. 1994). However, we did not detect this pig- in oligotrophic waters. From our results, the narrow ment in either Ochromonas isolate, consistent with its range of environmental conditions that permit absence in other chrysophytes (Lichtle et al. 1995, growth of the obligate mixotroph might seem disad- Tanabe et al. 2011). Differences in VAZ pool size vantageous, but this strategy is likely efficient in might underlie the higher NPQ capacity in stable environments with low-resource availability. CCMP2951, but cannot fully explain the large differ- Additionally, while the facultative mixotroph still ence between isolates. These might be due to other reached a higher percentage of its maximum components involved in NPQ such as the ancient growth rate under light limitation, this advantage CONTRASTING MIXOTROPHIC LIFESTYLES 65 might disappear under low but stable availability of light and prey, when close integration of photosyn- REFERENCES thesis and heterotrophy in the obligate mixotroph Allen, A. E., LaRoche, J., Maheswari, U., Lommer, M., Schauer, N., presumably preserves resources. Moreover, export of Lopez, P. J., Finazzi, G., Fernie, A. R. & Bowler, C. 2008. Whole- photosynthetically derived electrons to the mito- cell response of the pennate diatom Phaeodactylum tricornutum to – chondrial respiratory chain, even under low-light iron starvation. Proc.Natl.Acad.Sci.USA105:10438 43. Andersen, K. H., Aksnes, D. L., Berge, T., Fiksen, Ø. & Visser, A. conditions, could reduce the need to respire 2015. Modelling emergent trophic strategies in plankton. J. ingested carbon and therefore allow higher carbon Plankton Res. 37:862–8. assimilation efficiencies. Obligate mixotrophy could Andersson, A., Falk, S., Samuelsson, G. & Hagstrom,€ A. 1989. hence represent a specialization for stable resource- Nutritional characteristics of a mixotrophic nanoflagellate, Ochromonas sp. Microb. Ecol. 17:251–62. limited environments such as the oligotrophic Arsalane, W., Rousseau, B. & Duval, J. 1994. Influence of the pool oceans, while facultative mixotrophy would be more size of the xanthophyll cycle on the effects of light stress in a advantageous under fluctuating conditions as found diatom: Competition between photoprotection and photoin- in shallower coastal regions. hibition. Photochem. Photobiol. 60:237–43. The opposite responses to environmental drivers Azam, F., Fenchel, T., Field, J. G., Gray, J. S., Meyerreil, L. A. & Thingstad, F. 1983. The ecological role of water-column observed in the two marine Ochromonas species stud- microbes in the sea. Mar. Ecol. Prog. Ser. 10:257–63. ied here suggest that even closely related mixo- Bailey, S., Melis, A., Mackey, K. R. M., Cardol, P., Finazzi, G., van trophic protists will respond differently to changing Dijken, G., Berg, G. M., Arrigo, K., Shrager, J. & Grossman, ocean conditions with differing consequences for A. 2008. Alternative photosynthetic electron flow to oxygen in marine Synechococcus. Biochim. Biophys. Acta BBA Bioenerg. carbon cycling through marine microbial food webs. 1777:269–76. For instance, the nutritional balance of mixotrophs Bailleul, B., Berne, N., Murik, O., Petroutsos, D., Prihoda, J., might directly be affected by warming with a shift Tanaka, A., Villanova, V., et al. 2015. Energetic coupling toward a more heterotrophic nutrition at higher between plastids and mitochondria drives CO2 assimilation – temperatures (Wilken et al. 2013, Worden et al. in diatoms. Nature 524:366 9. Bailleul, B., Rogato, A., de Martino, A., Coesel, S., Cardol, P., 2015). Such a shift would be more likely in faculta- Bowler, C., Falciatore, A. & Finazzi, G. 2010. An atypical tive mixotrophs that have greater nutritional flexi- member of the light-harvesting complex stress-related pro- bility than in obligate mixotrophs which have a tein family modulates diatom responses to light. Proc. Natl. close interdependency of photosynthesis and Acad. Sci. USA 107:18214–9. Berge, T., Chakraborty, S., Hansen, P. J. & Andersen, K. H. 2017. heterotrophy. Further, expansion of oligotrophic Modeling succession of key resource-harvesting traits of mix- regions as oceans warm (Polovina et al. 2008) may otrophic plankton. ISME J. 11:212–23. provide conditions favoring the resource efficient Bird, D. F. & Kalff, J. 1986. Bacterial grazing by planktonic lake strategy, i.e., obligate mixotrophy. Understanding algae. Science 231:493–5. the physiological basis of different mixotrophic Boenigk, J., Pfandl, K., Stadler, P. & Chatzinotas, A. 2005. High diversity of the ‘Spumella-like’ flagellates – an investigation strategies in a broader assortment of environmentally based on the SSU rRNA gene sequences of isolates from relevant models will help to predict their respective habitats located in six different geographic regions. Environ. biogeography and contributions to food web dynam- Microbiol. 7:685–97. ics and carbon cycling in the future oceans. del Campo, J. & Massana, R. 2011. Emerging diversity within chrysophytes, choanoflagellates and bicosoecids based on molecular surveys. Protist 162:435–48. ACKNOWLEDGMENTS Caron, D. A., Porter, K. G. & Sanders, R. W. 1990. Carbon, nitro- gen, and phosphorus budgets for the mixotrophic phytoflag- ellate Poterioochromonas malhamensis (chrysophyceae) during – We thank E. Ruby for providing the Vibrio fischeri strain, F. bacterial ingestion. Limnol. Oceanogr. 35:433 43. Chavez for access to a radioactive isotope lab, T. Pennington Caron, D. A., Sanders, R. W., Lim, E. L., Marrase, C., Amaral, L. for help with establishing the carbon fixation protocol, D. Kli- A., Whitney, S., Aoki, R. B. & Porter, K. G. 1993. Light- dependent phagotrophy in the freshwater mixotrophic chrys- mov for help with FRRf-setup, and M. Hamilton for help with ophyte Dinobryon cylindricum. Microb. Ecol. 25:93–111. sampling. Finally, we thank R. Massana for cosponsoring SW’s Castresana, J. 2000. Selection of conserved blocks from multiple Marie Curie Postdoctoral Fellowship. This work was sup- alignments for their use in phylogenetic analysis. Mol. Biol. ported by grants from the Department of Energy (DOE-DE- Evol. 17:540–52. SC0004765) and Gordon and Betty Moore Foundation Cuvelier, M. L., Allen, A. E., Monier, A., McCrow, J. P., Messie, M., (GBMF3307 and GBMF3788) to AZW and SW was supported Tringe, S. G., Woyke, T. et al. 2010. Targeted metagenomics largely by a Postdoctoral Fellowship from the European and ecology of globally important uncultured eukaryotic phy- Union’s Seventh Framework Programme under the Marie toplankton. Proc. Natl. Acad. Sci. USA 107:14679–84. Skłodowska-Curie grant agreement No 626182. Cuvelier, M. L., Guo, J., Ortiz, A. C., van Baren, M. J., Akram Tariq, M., Partensky, F. & Worden, A. Z. 2017. Responses of the picoprasinophyte Micromonas commoda to light and ultra- AUTHOR CONTRIBUTIONS violet stress. PLoS ONE 12:e0172135. Dunn, A. K., Millikan, D. S., Adin, D. M., Bose, J. L. & Stabb, E. AZW and SW planned and designed the research; V. 2006. New rfp- and pES213-derived tools for analyzing SW performed experiments and analyzed physiology symbiotic Vibrio fischeri reveal patterns of infection and lux data; and CJC performed sequencing and phyloge- expression in situ. Appl. Environ. Microbiol. 72:802–10. netic analysis. SW drafted the manuscript with input Estep, K. W., Davis, P. G., Keller, M. D. & Sieburth, J. M. 1986. How important are oceanic algal nanoflagellates in bac- from AZW and additional edits from CJC. terivory? Limnol. Oceanogr. 31:646–50. 66 SUSANNE WILKEN ET AL.

Field, C. B., Behrenfeld, M. J., Randerson, J. T. & Falkowski, P. Microbial Eukaryote Transcriptome Sequencing Project 1998. Primary production of the biosphere: Integrating ter- (MMETSP): illuminating the functional diversity of eukary- restrial and oceanic components. Science 281:237–40. otic life in the oceans through transcriptome sequencing. Fischer, R., Giebel, H. A., Hillebrand, H. & Ptacnik, R. 2017. PLoS Biol. 12:e1001889. Importance of mixotrophic bacterivory can be predicted by Keller, M. D., Selvin, R. C., Claus, W. & Guillard, R. R. L. 1987. light and loss rates. Oikos 126:713–22. Media for the culture of oceanic ultraphytoplankton. J. Phy- Floder,€ S., Hansen, T. & Ptacnik, R. 2006. Energy-dependent bac- col. 23:633–8. terivory in Ochromonas minima – a strategy promoting the use Kimura, B. & Ishida, Y. 1989. Phospholipid as a growth factor of of substitutable resources and survival at insufficient light Uroglena americana, a red tide chrysophyceae in Lake Biwa. supply. Protist 157:291–302. Nippon Suisan Gakkaishi 55:799–804. Flynn, K. J. & Mitra, A. 2009. Building the “perfect beast”: mod- Kirkham, A. R., Lepere, C., Jardillier, L. E., Not, F., Bouman, H., elling mixotrophic plankton. J. Plankton Res. 31:965–92. Mead, A. & Scanlan, D. J. 2013. A global perspective on mar- Flynn, K. J., Stoecker, D. K., Mitra, A., Raven, J. A., Glibert, P. M., ine photosynthetic picoeukaryote community structure. ISME Hansen, P. J., Graneli, E. & Burkholder, J. M. 2013. Misuse J. 7:922–36. of the phytoplanktonzooplankton dichotomy: the need to Kolber, Z. S., Prasil, O. & Falkowski, P. G. 1998. Measurements assign organisms as mixotrophs within plankton functional of variable chlorophyll fluorescence using fast repeti- types. J. Plankton Res. 35:3–11. tion rate techniques: Defining methodology and experi- Graf, J., Dunlap, P. V. & Ruby, E. G. 1994. Effect of transposon- mental protocols. Biochim. Biophys. Acta BBA Bioenerg. induced motility mutations on colonization of the host light 1367:88–106. organ by Vibrio fischeri. J. Bacteriol. 176:6986–91. Lane, C. E. & Archibald, J. M. 2008. The eukaryotic tree of life: Grossmann, L., Bock, C., Schweikert, M. & Boenigk, J. 2016. endosymbiosis takes its TOL. Trends Ecol. Evol. 23:268–75. Small but manifold – hidden diversity in ‘Spumella-like flagel- Leles, S. G., Mitra, A., Flynn, K. J., Tillmann, U., Stoecker, D., lates’. J. Euk. Microbiol. 63:419–39. Jeong, H. J., Burkholder, J. et al. 2019. Sampling bias misrep- Guindon, S., Dufayard, J. F., Lefort, V., Anisimova, M., Hordijk, resents the biogeographical significance of constitutive mixo- W. & Gascuel, O. 2010. New algorithms and methods to esti- trophs across the global oceans. Global Ecol. Biogeogr. 28:418– mate maximum-likelihood phylogenies: assessing the perfor- 28. mance of PhyML 3.0. Systematic Biol. 59:307–21. Leles, S. G., Polimene, L., Bruggeman, J., Blackford, J., Ciavatta, Guo, J., Wilken, S., Jimenez, V., Choi, C. J., Ansong, C., Dan- S., Mitra, A. & Flynn, K. J. 2018. Modelling mixotrophic nebaum, R., Sudek, L. et al. 2018. Specialized proteomic functional diversity and implications for ecosystem function. responses and an ancient photoprotection mechanism sus- J. Plankton Res. 40:627–42. tain marine green algal growth during phosphate limitation. Lepere, C., Vaulot, D. & Scanlan, D. J. 2009. Photosynthetic Nature Microbiol. 3:781–90. picoeukaryote community structure in the South East Pacific Hansen, P. J. 2011. The role of photosynthesis and food uptake Ocean encompassing the most oligotrophic waters on Earth. for the growth of marine mixotrophic dinoflagellates. J. Environ. Microbiol. 11:3105–17. Eukaryot. Microbiol. 58:203–14. Lichtle, C., Arsalane, W., Duval, J. C. & Passaquet, C. 1995. Char- Hansen, P. J. & Hjorth, M. 2002. Growth and grazing responses of acterization of the light-harvesting complex of Giraudyopsis Chrysochromulina ericina (Prymnesiophyceae): the role of irradi- stellifer (Chrysophyceae) and effects of light stress. J. Phycol. ance, prey concentration and pH. Mar. Biol. 141:975–83. 31:380–7. Hartmann, M., Grob, C., Tarran, G. A., Martin, A. P., Burkill, P. Lie, A. A. Y., Liu, Z., Terrado, R., Tatters, A. O., Heidelberg, K. H., Scanlan, D. J. & Zubkov, M. V. 2012. Mixotrophic basis B. & Caron, D. A. 2018. A tale of two mixotrophic chryso- of atlantic oligotrophic ecosystems. Proc. Natl. Acad. Sci. USA phytes – insights into the metabolisms of two Ochromonas spe- 109:5756–60. cies (Chrysophyceae) through a comparison of gene Hartmann, M., Zubkov, M. V., Scanlan, D. J. & Lepere, C. 2013. expression. PLoS ONE 13:e0192439. In situ interactions between photosynthetic picoeukaryotes MacIntyre, H. L., Kana, T. M., Anning, T. & Geider, R. J. 2002. and bacterioplankton in the atlantic ocean: evidence for Photoacclimation of photosynthesis irradiance response mixotrophy. Environ. Microbiol. Reports 5:835–40. curves and photosynthetic pigments in microalgae and Havskum, H. & Hansen, A. S. 1997. Importance of pigmented cyanobacteria. J. Phycol. 38:17–38. and colourless nano-sized protists as grazers on nanoplank- McDonald, S.M., Plant, J.N. & Worden, A.Z. 2010. The mixed lin- ton in a phosphate-depleted Norwegian fjord and in enclo- eage nature of nitrogen transport and assimilation in marine sures. Aquat. Microb. Ecol. 12:139–51. eukaryotic phytoplankton: a case study of Micromonas. Mol. Heinbokel, J. F. 1978. Studies on the functional role of tintinnids Biol. Evol. 27: 2268-83. in the southern California bight I. Grazing and growth rates Milligan, A. J., Halsey, K. H. & Behrenfeld, M. J. 2015. Advancing in laboratory cultures. Mar. Biol. 47:177–89. interpretations of 14C-uptake measurements in the context Jakob, T., Goss, R. & Wilhelm, C. 1999. Activation of diadinoxan- of phytoplankton physiology and ecology. J. Plankton Res. thin de-epoxidase due to a chlororespiratory proton gradient 37:692–8. in the dark in the diatom Phaeodactylum tricornutum. Plant Millikan, D. S. & Ruby, E. G. 2003. FlrA, a r54-dependent tran- Biol. 1:76–82. scriptional activator in Vibrio fischeri, is required for motility Jardillier, L., Zubkov, M. V., Pearman, J. & Scanlan, D. J. 2010. and symbiotic light-organ colonization. J. Bacteriol. 185:3547– Significant CO2 fixation by small prymnesiophytes in the sub- 57. tropical and tropical northeast Atlantic ocean. ISME J. Mitra, A., Flynn, K. J., Burkholder, J. M., Berge, T., Calbet, A., 4:1180–92. Raven, J. A., Graneli, E. et al. 2014. The role of mixotrophic Jeong, H. J., Yoo, Y. D., Kim, J. S., Seong, K. A., Kang, N. S. & protists in the biological carbon pump. Biogeosciences 11:995– Kim, T. H. 2010. Growth, feeding and ecological roles of 1005. mixotrophic and heterotrophic dinoflagellates in marine Mitra, A., Flynn, K. J., Tillmann, U., Raven, J. A., Caron, D., planktonic food webs. Ocean Sci. J. 45:65–91. Stoecker, D. K., Not, F. et al. 2016. Defining planktonic pro- Katoh, K. & Standley, D. M. 2013. MAFFT multiple sequence tist functional groups on mechanisms for energy and nutri- alignment software version 7: improvements in performance ent acquisition - incorporation of diverse mixotrophic and usability. Mol. Biol. Evol. 30:772–80. strategies. Protist 167:106–20. Keeling, P. J. 2013. The number, speed, and impact of pastid Moon-van der Staay, S. Y., De Wachter, R. & Vaulot, D. 2000. endosymbioses in eukaryotic evolution. Annu. Rev. Plant Biol. Oceanic 18S rDNA sequences from picoplankton reveal 64:583–607. unsuspected eukaryotic diversity. Nature 409:607–10. Keeling, P. J., Burki, F., Wilcox, H. M., Allam, B., Allen, E. E., Orsi, W. D., Wilken, S., del Campo, J., Heger, T., James, E., Amaral-Zettler, L. A., Armbrust, E. V. et al. 2014. The Marine Richards, T. A., Keeling, P. J., Worden, A. Z. & Santoro, A. CONTRASTING MIXOTROPHIC LIFESTYLES 67

E. 2018. Identifying protist consumers of photosynthetic Unrein, F., Gasol, J. M., Not, F., Forn, I. & Massana, R. 2014. Mix- picoeukaryotes in the surface ocean using stable isotope otrophic haptophytes are key bacterial grazers in olig- probing. Environ. Microbiol. 20:815–27. otrophic coastal waters. ISME J. 8:164–76. Palsson, C. & Daniel, C. 2004. Effects of prey abundance and Unrein, F., Massana, R., Alonso-Saez, L. & Gasol, J. M. 2007. Sig- light intensity on nutrition of a mixotrophic flagellate and nificant year-round effect of small mixotrophic flagellates on its competitive relationship with an obligate heterotroph. bacterioplankton in an oligotrophic coastal system. Limnol. Aquat. Microb. Ecol. 36:247–56. Oceanogr. 52:456–69. Peers, G., Truong, T. B., Ostendorf, E., Busch, A., Elrad, D., Van Heukelem, L. & Thomas, C. S. 2001. Computer-assisted high- Grossman, A. R., Hippler, M. & Niyogi, K. K. 2009. An performance liquid chromatography method development ancient light-harvesting protein is critical for the regulation with applications to the isolation and analysis of phytoplank- of algal photosynthesis. Nature 462:518–21. ton pigments. J. Chromatogr. A 910:31–49. Pennington, J. T. & Chavez, F. P. 2000. Seasonal fluctuations of Vanlerberghe, G. C. & McIntosh, L. 1997. Alternative oxidase: temperature, salinity, nitrate, chlorophyll and primary pro- From gene to function. Annu. Rev. Plant Physiol. Plant Mol. duction at station H3/M1 over 1989–1996 in Monterey bay, Biol. 48:703–34. California. Deep Sea Res. Part 2 Top. Stud. Oceanogr. 47:947–73. Ward, B. A. & Follows, M. J. 2016. Marine mixotrophy increases Polovina, J. J., Howell, E. A. & Abecassis, M. 2008. Ocean’s least trophic transfer efficiency, mean organism size, and vertical productive waters are expanding. Geophys. Res. Lett. 35:L03618. carbon flux. Proc. Natl. Acad. Sci. USA 113:2958–63. Pomeroy, L. R. 1974. The ocean’s food web, a changing para- Webb, W. L., Newton, M. & Starr, D. 1974. Carbon dioxide digm. Bioscience 24:499–504. exchange of Alnus rubra. Oecologia 17:281–91. Pruesse,E.,Quast,C.,Knittel,K.,Fuchs,B.M.,Ludwig,W.,Peplies,J. Wilken, S., Huisman, J., Naus-Wiezer, S. & Van Donk, E. 2013. & Glockner, F. O. 2007. SILVA: a comprehensive online Mixotrophic organisms become more heterotrophic with ris- resource for quality checked and aligned ribosomal RNA ing temperature. Ecology Lett. 16:225–33. sequence data compatible with ARB. Nucl. Acids Res. 35:7188–96. Wilken, S., Schuurmans, J. M. & Matthijs, H. C. P. 2014a. Do mix- Ronquist, F., Teslenko, M., van der Mark, P., Ayres, D. L., Dar- otrophs grow as photoheterotrophs? photophysiological ling, A., Hohna, S., Larget, B., Liu, L., Suchard, M. A. & acclimation of the chrysophyte Ochromonas danica after feed- Huelsenbeck, J. P. 2012. MrBayes 3.2: efficient Bayesian phy- ing. New Phytol. 204:882–9. logenetic inference and model choice across a large model Wilken, S., Verspagen, J. M. H., Naus-Wiezer, S., Van Donk, E. & space. Systematic Biol. 61:539–42. Huisman, J. 2014b. Comparison of predator-prey interactions Rothhaupt, K. O. 1996. Utilization of substitutable carbon and with and without intraguild predation by manipulation of phosphorus sources by the mixotrophic chrysophyte Ochromo- the nitrogen source. Oikos 123:423–32. nas sp. Ecology 77:706–15. Worden, A. Z. 2006. Picoeukaryote diversity in coastal waters of Rothhaupt, K. O. 1997. Nutrient turnover by freshwater bacterivo- the Pacific Ocean. Aquat. Microb. Ecol. 43:165–75. rous flagellates: Differences between a heterotrophic and a Worden, A. Z., Follows, M. J., Giovannoni, S. J., Wilken, S., Zim- mixotrophic chrysophyte. Aquat. Microb. Ecol. 12:65–70. merman, A. E. & Keeling, P. J. 2015. Rethinking the marine Sanders, R. W., Berninger, U. G., Lim, E. L., Kemp, P. F. & carbon cycle: Factoring in the multifarious lifestyles of Caron, D. A. 2000. Heterotrophic and mixotrophic microbes. Science 347:1257594. nanoplankton predation on picoplankton in the Sargasso Worden, A. Z., Lee, J. H., Mock, T., Rouze, P., Simmons, M. P., Sea and on George’s Bank. Mar. Ecol. Prog. Ser. 192:103–18. Aerts, A. L., Allen, A. E. et al. 2009. Green evolution and Sanders, R. W., Porter, K. G. & Caron, D. A. 1990. Relationship between dynamic adaptations revealed by genomes of the marine phototrophy and phagotrophy in the mixotrophic chrysophyte picoeukaryotes Micromonas. Science 324:268–72. Poterioochromonas malhamensis. Microb. Ecol. 19:97–109. Worden, A. Z., Seidel, M., Smriga, S., Wick, A., Malfatti, M., Bar- Schuback, N., Schallenberg, C., Duckham, C., Maldonado, M. T. tlett, D. & Azam, F. 2006. Trophic regulation of Vibrio cholera & Tortell, P. D. 2015. Interacting effects of light and iron in coastal marine waters. Environ. Microbiol. 8:21–9. availability on the coupling of photosynthetic electron trans- port and CO2-assimilation in marine phytoplankton. PLoS ONE 10:e0133235. Selosse, M. A., Charpin, M. & Not, F. 2017. Mixotrophy every- Supporting Information where on land an in water: the grand ecart hypothesis. Ecol. Let. 20:246–63. Additional Supporting Information may be Stabb, E. V., Reich, K. A. & Ruby, E. G. 2001. Vibrio fischeri genes hvnA and hvnB encode secreted NAD(+)-glycohydrolases. J. found in the online version of this article at the Bacteriol. 183:309–17. publisher’s web site: Stamatakis, A. 2014. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics Figure S1. Abundances of Ochromonas and the 30:1312–3. bacterial prey Vibrio fischeri in semi-continuous cul- Stoecker, D. K. 1998. Conceptual models of mixotrophy in plank- tures of the two Ochromonas isolates with or with- tonic protists and some ecological and evolutionary implica- out daily prey-amendments. tions. Eur. J. Protistol. 34:281–90. Stoecker, D. K. & Michaels, A. E. 1991. Respiration, photosynthe- Figure S2. Forward-angle light scatter (FALS) of sis and carbon metabolism in planktonic ciliates. Mar. Biol. Ochromonas isolates under prey-amended and 108:441–7. Tanabe, Y., Shitara, T., Kashino, Y., Hara, Y. & Kudoh, S. 2011. prey-deplete conditions. Utilizing the effective xanthophyll cycle for blooming of Figure S3. Molar ratio of diadinoxanthin con- Ochromonas smithii and O. itoi (Chrysophyceae) on the snow surface. PLoS ONE 6:e14690. tent relative to chlorophyll a in the two Ochromo- Tittel, J., Bissinger, V., Zippel, B., Gaedke, U., Bell, E., Lorke, A. nas isolates under prey-amended and prey-deplete & Kamjunke, N. 2003. Mixotrophs combine resource use to conditions. outcompete specialists - implications for aquatic food webs. Proc. Natl. Acad. Sci. USA 100:12776–81.